[Prev][Next][Index][Thread]
Re: ANN: "Interacting with CORBA" article
On Thu, 23 Dec 1999 21:23:35 GMT, Michi Henning <michi@ooc.com.au> wrote:
> On Thu, 23 Dec 1999, Jason Trenouth wrote:
>
> > But that is beside the point, which is: why did
> > IDLScript itself need a new kind of specification process? Why isn't it simply
> > another language binding?
>
> I remember that there was considerable discussion about precisely this question
> when the RFP was drafted. I didn't follow all the details back then but,
> from memory, I think there was a faction who felt that CORBA needed a
> scripting language, but that it shouldn't go and favor any specific existing
> one. (If you pick Perl, you alienate the Tcl crowd, if you pick Tcl, you
> alienate the Python crowd, etc...)
So why raise any of them above the others?
> I suspect that, from a vendor's perspective, a single scripting language
> is also more attractive than a whole raft of language bindings. At least,
> with a single scripting language, there is only one thing to support and
> maintain; on the other hand, supporting a binding to Tcl, and another one
> to Perl, and another one to Python, etc, requires more work (and
> integration with public domain or open source software that may keep
> changing independently).
Except that this reasoning breaks down since two appear to have been chosen.
So the reason for the RFP was to encourage ORB vendors to support a standard
scripting language in addition to the 'proper' languages they were already
supporting? This is so that ORB customers can all write in the same scripting
language even though they aren't all writing in the same 'proper' language?
What's the point of that?
__Jason
Follow-Ups:
References: